Automation in Construction 10 2001.

507516
www.elsevier.comrlocaterautcon

Automating utility route design and planning through GIS

Min-Yuan Cheng ) , Guey-Lin Chang
Department of Construction Engineering, National Taiwan Uniersity of Science and Technology, P.O. Box 90 130, Taipei, Taiwan
Accepted 30 August 2000

Abstract
In trench construction, one of the tasks for engineers is to select a suitable route to minimize construction cost and
obstructions. This paper discusses the development of a Geographic Information Systems GIS.-based system to automate
the process of routing and design of an underground power supply system. In the system, surface and underground utilities
are represented in several coverages. Using network analysis, the system determines the optimal paths for routing the
utilities. Through database queries and spatial operations, the construction conflict points between the basic coverages and
the selected route are not only identified, but a reallocation schedule is also determined. q 2001 Elsevier Science B.V. All
rights reserved.
Keywords: GIS; Automating; Utility route

1. Introduction
Unlike most civil engineering projects, trenching
schemes are not confined within a recognized site
but by their nature, lie between widely spread out
points of connection and can involve working in a
variety of rural and urban environments. For most
electrical supply system tunnel projects, construction
sites are often in close contact with existing public
utility lines and structures. The underground apparatuses of most public utilities include live systems and
principal services, thus, water and gas can be dangerous to trench operatives when damaged or fractured.
Especially in urban areas, obstacles such as existing
)
Corresponding author. Tel.: q886-2-27376596; fax: q886-227376606.
E-mail address: myc@hp.ct.ntust.edu.tw M.-Y. Cheng..

underground public utility lines, railways, canals,

and roads can influence the route significantly as
there may only be a limited number of feasible
crossing points. Thus, selecting a suitable route to
avoid existing obstacles not only reduces the risk of
damaging existing utilities, but also minimizes the
cost and duration required for construction.
The primary drawback of current planning methods is that the information required for trench route
planning and design is incomplete and stored in
different administrative authorities of the city government. Currently, the existing public utility historical data describing pipe distributions, pipe types, and
pipe dimensions, is not well recorded and managed.
In addition, as various types of utilities are managed
by different city government authorities, the coordination between these utility administrations is often
lacking. Thus, for a trench project, the information

0926-5805r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 6 - 5 8 0 5 0 0 . 0 0 1 0 3 - 5

508

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

required for route planning is limited and the designer has to accomplish his work based on incomplete information. In the construction phase, due to
the lack of accurate existing underground utility
information, the constructor is apt to damage and
fracture existing utilities and cause construction delay. The conventional approaches of electricity supply route planning rely on heuristic strategies and
rules of thumb based on an individuals knowledge
and experience gained from previous problem solutions. The complexity of the problem involves the
need to process spatial data such as road and circuit
maps, public utility records, and construction plans.
In the classic manual approach, it is difficult for the
planner to identify the spatial relationship between
different data layers which usually results in an
overdesigned, costly and non-standardized solution.
Summarizing the drawbacks of the current approach for electricity route planning and design,
Geographic Information Systems GIS., a technology
for spatial data analysis, seems to have significant
potential in solving these problems. By using GIS,
planners can integrate and organize ill-structured
construction information such as circuit maps and
records, road maps and associated attributes, construction maps, and other spatial information more
efficiently. Due to its spatial operation capability,
GIS helps planners identify the spatial relationships
between different data layers. In addition, GIS not
only stores the spatial features and attributes, but can
also be used as an analysis tool. GIS-ArcrInfo provides an internal macro language for users to automate their system design and the ability to interface
with other application softwares. Hence, heuristic
methods and algorithms can be developed to solve
utility route planning and design as an ill-structured
problem.

and dimensions for the underground utilities to minimize the influence on existing utility lines and structures; 3. identify the conflict points between the
designed route and the existing utilities; 4. integrate
construction schedule and design information to create a reallocation schedule for existing public utilities; and 5. provide distribution information for the
existing underground utilities to reduce potential
damages or fractures caused by construction excavation.

2. Research objectives

4.1.2. Rules for route planning

Depending on the designers experience, regulation restrictions and the complexity of the problem,
determination of the optimal route may vary from
project to project. Most of the route designs are
based on an individuals experience and common
sense gained from previous projects. However, there
still exist some rule-of-thumb practices in construc-

The primary purpose of this paper is to investigate

the potential for automating the underground public
utility system routing and design process. The objectives required to achieve this primary purpose are the
following: 1. identify the routing alternatives along
existing streets; 2. determine the optimal position

3. Scope definition
The system, RoutePlan, is developed to assist
engineers in selecting a suitable route to minimize
construction cost and obstructions. In this study, the
system is mainly focused on route planning and
design of an electric utility line, and the open trench
method is the construction method used in excavation. However, using the same methodology and
algorithms with appropriate modifications, RoutePlan
can be applied to a variety of different public utility
design and construction methods.

4. System development
4.1. Knowledge acquisition and data preparation
4.1.1. Knowledge resources
This study reviewed the field practice and identified contemporary state of art route planning. To
obtain the experts knowledge and experience in
power supply route design, four resources were compiled and reviewed: 1. interviews with experts, 2.
previous research, 3. design handbooks, and 4.
standard construction practice.

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

tion management for carrying out route planning.

Therefore, in this study, knowledge and experience
for route planning was compiled into two categories:
1. primary rules and 2. specific concerns. Primary
rules are the general rules and regulations that have
to be applied to most utility system routing and
design, while specific concerns include the unique
aspect of trench excavation in urban areas. The
primary rules, including the minimum road width
requirement, routing through private lands, unstable
underground geotechnical conditions, canals, underpasses, artifact remains, and government restriction
areas, etc., are the criteria used to select the potential
routes that satisfy the regulation restrictions. The
specific concerns include five categories: 1. traffic
low interference, 2. pavement materials, 3. planned
roads, 4. future planned utility conduits, and 5.
existing utility tunnels; they are the factors used to
select and determine the optimal route.
4.1.3. AHP analysis
Specific concern factors are the fundamentals used
for selecting the optimal route. Fig. 1 shows the
hierarchical structure of influencing factors. In order
to obtain a quantitative result for the relative weight

509

of factors objectively, a questionnaire survey was

conducted using the AHP method. Forty copies of
the questionnaires were sent out and 24 effective
copies were returned showing a recovery rate of
60%. The effective samples represent a variety of
field professionals, including Taiwan Power Co. designers, construction managers, field supervisors,
transportation experts and professors totaling 24 people. In this survey, based around personal experiences and understanding of route planning and design, all the factors weights, shown in Fig. 1, were
determined on a pair-wise comparison basis. All
questionnaires underwent a contingency test to ensure that they were logically correct. The data is used
as the basis for quantitative evaluation of the optimal
route in Section 4.1.4.
4.1.4. Impedance index analysis to identify the optimal route
Impedance Index II. analysis is an objective
function used in this study to rank the potential
routes and determine the optimal choice. Based on
the AHP analysis results, II is calculated by summing the product of the length and relative weighting
of the road sections. For each potential route, the II

Fig. 1. Hierarchy structure of influencing factors for route selection.

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

510

can be described and calculated using Eqs. 1. 4.,

as shown below.
n

Tk s

Wi

1.

is1

IVk s D k =

1
Tk

2.

II R j s

IVk

3.

ks1

Objective function:
II s Min II R j . ,

4.

where Tk : total weight of road section k; Wi : relative

weight of the specific concern factor i; n: number of
specific concern factors; IVk : impedance value of
road section k; Dk : length of road section k; II R j :
impedance index of the potential route j; s: number
of road sections for each potential route; II:
impedance index.
Each potential route is ranked based on the II R j
see Eq. 3... As shown in Eq. 4., the potential route
with the minimal II is the optimal route.
4.2. System functions
The functional modules of the system are developed according to the capabilities and information
needed for the user to conduct route planning and
design. The primary functions of the system are
described as follows.
Graphical display: The graphical display is a
graphic representation of the site features and the
associated tabular attributes. It is a communication
and functions as a decision support system which
assists the user to identify the distributions of the
existing public utility lines, conflict points, and the
optimal route. Depending on the user needs, the site
coverages and analysis results can be displayed and
overlaid on the screen in either a detailed or an
abstract form. The querying of the tabular attributes
is the function used to display the labels associated
with the site coverage and the integrated information
of a particular entity in the coverage.
Database query: The related databases of site
coverages can be identified using the database query.
v

The main difference between database query and

graphical display is that the database query can
manipulate and list the integrated coverage information without displaying the site features. For example, without opening the road coverage, the user can
query and list its attributes such as road width,
number of lanes, lane width, road condition, traffic
condition, etc.
Spatial analysis: The spatial analysis module is
the function used to identify the spatial relationships
between the site features and the potential routes.
The six ArcrInfos OVERLAY functions, namely
Reselect, Buffer, Erasecov, Union, Intersect and
Identity, provided in this module allow the user to
conduct route planning manually.
Route planning: The Route Planning module is
a function that allows the user to identify a suitable
route for design. Four primary functions, viz., potential route analysis, optimal route selection, locate
conflict points, and reallocation schedule, are provided to achieve the task of route planning and
design.
v

4.3. System design and deelopment algorithm

RoutePlan consists of knowledge and databases
specific to the utility route planning and algorithms
for integrating and automating the design. The system, as shown in Fig. 2, is an integration of GIS with
database management systems. The system algorithm
is primary divided into four parts, viz., potential
route analysis, determination of the optimal route,
location of the conflict points, and identification of
the reallocation schedule for existing utilities. The
system design flow is discussed as follows.
4.3.1. Potential routes analysis
This section discusses the process of analysis for
potential routes. An ArcrInfo dictionary, including a
list of coverages, the names of attributes, and a
description of the attribute values, is created for
reference during development of the ArcrInfo project database. After creating the site geometry and
modeling the coverages, the concept of Searching by
Elimination is used to develop an algorithm for
analyzing the potential routes between two nodes in

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

511

Fig. 2. System design and development algorithm.

the road network see Fig. 3.. Considering the constraints and selection criteria, routes which do not

meet the requirements are eliminated from the network.

512

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

Fig. 3. Potential route analysis.

4.3.2. Determining the optimal route

The primary function of optimal route selection is
to identify the route of least impedance between two
nodes in the road network. The systems route selec-

tion process is developed using ArcrInfos Network

functions. Initially, the dominant factors and weight
of factors are identified and added to the road coverage related database. Then, an objective function

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

called impedance index II. is developed to determine the optimal route. The II is calculated by
summing the product of the length and relative
weighting of the road sections. The route with least
impedance between two nodes is identified as the
optimal route.
4.3.3. Locating conflict points
This section is to identify the conflict points
between the selected route and the existing public
utilities see Fig. 4.. The selected route is first
buffered based on the width of excavation. The
existing facilities such as underground utility lines,
railways, canals, motorways, etc., within the excavation depth are selected and represented as a coverage. Using the ArcrInfos Overlay functions, the

513

conflict points between the selected route and the

existing facilities are then identified.
4.3.4. Creating a reallocation schedule for existing
utilities
Based on the construction progress and distances
from the start node of the selected route to the
identified conflict points, a reallocation schedule for
existing utilities is created. In the system, an Excel
macro is used to calculate the distance between two
nodes and determine a schedule for reallocation.
4.4. System architecture
The architecture of the system involves identifying the tools used to develop the system functional

Fig. 4. Process for locating conflict points.

514

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

Fig. 5. System architecture.

modules and the means by which each interfaces

with one another and the user. Fig. 5 shows the
system architecture. The user and program interface
for RoutePlan is established at three levels: applica-

tion user interface, command user interface, and

program data interface.
The application user interface was built on top of
the command user interface, using the Simple Macro

Fig. 6. Result of potential route analysis.

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

515

Fig. 7. Analysis results of conflict points and reallocation schedule.

Language SML.. A set of the application user interface objects, including pull-down menus, pop-up
menus and forms, was developed for the system.
These objects were designed for the specific application and provide a non-command-driven user interface. The command user interface is a simple command language, similar in purpose to operating system command languages found in Unix, MS-DOS,
and other operating systems. In addition to automating the system design and directing the program flow
of control, the command user interface is also used
to integrate ArcView and the Excel spreadsheet. The
program data interface writesrreads the information
torfrom the associated databases. In this way, the
data files which are stored in a standard dBASE file
format act as the communication media.
The systems prime components include Visual
Basic, CAD, ArcrInfo, ArcView, Excel spreadsheet,
and wARCx prompt. The user communicates with the
system components through a custom interface developed in Visual Basic. The CAD software AutoCAD is used to create the site geometry. Using the
ArcrInfo functional modules such as Starter Kit,
Arcplot, Overlay, Tables, and Network, the route
plan and design is mapped out. ArcView is used to
perform the graphical query and display the analysis
results. Excel spreadsheet is used to add additional
tabular attributes and calculate the impedance index.

To enhance flexibility, the system allows the user to

suspend execution of the SML macros and temporarily exit to the wARCx prompt to use the ArcrInfo
functions.
4.5. System demonstration
This study incorporated a Taipei City map with
1r5000 scale for system testing. Fig. 6 shows the
street map after potential route analysis. In the figure, the bold line represents the optimal route between the start and finish nodes for utility route
planning. The conflict points and reallocation schedule, as shown in Fig. 7, are also identified and
calculated. RoutePlan performs a series of spatial
operations and database queries to identify the optimal route and construction conflict points. In comparison with current methods, RoutePlan provides an
environment which saves significant computational
effort while increasing design accuracy and consistency.

5. Conclusions
This papers endeavor focuses on developing an
automated routing system to assist engineers in selecting a suitable route to avoid existing obstacles

516

M.-Y. Cheng, G.-L. Changr Automation in Construction 10 (2001) 507516

and minimize construction cost and scheduling. The

system, RoutePlan, systematically organizes incomplete and ill-structured information in the following
areas: 1. development of the route selection knowledge base KB. to represent the experts knowledge
and experience for route planning; 2. application of
the AHP method to identify the dominant factors for
the optimal route selection; and 3. development of
the ArcrInfo database to contain the existing utility
coverages and associated feature attribute tables.
RoutePlan improves route-planning efficiency
through the integration of locational and thematic
information into a single environment. This capability provided in ArcrInfo, not only speeds up the
planning process by avoiding data extraction from
various resources but more importantly, ensures data
accuracy and consistency.

Further reading
J.K. Budleigh, Trench Excavation and Support, Billing and Sons,
Worcester, UK, 1989.
M.Y. Cheng, J.T. OConnor, Site layout of construction temporary
facilities using enhanced-Geographical Information System
GIS., Autom. Constr. 3 1994. 1119.
M.Y. Cheng, Automated site layout of temporary construction
facilities using Geographic Information Systems GIS., PhD
Diss., Department of Civil Engineering, University of Texas,
1992.
M.E. Helander, E. Melachrinoudis, Facility location and reliable
route planning in hazardous material transportation, Transp.

Sci. 31 3. 1997. 216226, Aug., MD, USA, ISSN: 00411655.

M.Y. Cheng, K. Varghese, J.T. OConnor, Management of spatial
information for construction planning and design using Geographic Information Systems GIS., Proc. 9th Int. Symp.
Automn. Rob. Constr. 2 1992. 393402.
D. Stein, K. Mollers, R. Bielecki, Microtunnelling, Ernst and
Sohn, Birlin, 1989.
E.-C. Yeh, Z. Sumic, S.S. Venkata, APR: a Geographic Information System based primary router for underground residential
distribution design, IEEE Trans. Power Syst. 10 1. 1995.
400406, Feb., NJ, USA, ISSN: 0885-8950.
R. Greene, N. Agbenowosi, G.V. Loganathan, GIS-based approach to sewer system design, J. Surv. Eng. 125 1. 1999.
3657, Feb., ASCE, Reston, VA, USA, ISSN: 0733-9453.
Y. Saab, M. VanPutee, Shortest path planning on topographical
maps, IEEE Trans. Power Syst., Man, Cybern., Part A: Syst.
and Hum. 29 1. 1999. 139150, Jan., NJ, USA, ISSN:
1083-4427.
P. Modesti, A. Sciomachen, Utility measure for finding multiobjective shortest paths in urban multimodal transportation networks, Eur. J. Oper. Res. 111 3. 1998. 495508, Dec.,
Elsevier Sci., Amsterdam, Netherland, ISSN: 0377-2217.
S. Soehodho, Development of genetic-toxonomy evaluator for
finding shortest path in transportation planning systems, Proc.
of the 1997 3rd Conf. on Urban Transp. and Environ. for the
21 Cent., Sep., Acquasparta, Italy, Computational Mechanics
Pub., 1997, pp. 177191.
L.E. Kavraki, M.N. Kolountzakis, J.-C. Latombe, Analysis of
probabilistic roadmaps for path planning, IEEE Trans. Rob.
Autom. 14 1. 1998. 166171, Feb., NJ, USA, ISSN: 09189715.
Environmental Systems Research Institute, PC ArcrInfo Reference Manual, Redlands, CA.
K. Varghese, Automated route planning for large vehicles on
industrial construction sites, PhD Diss., Department of Civil
Engineering, University of Texas, 1992.